Retroviruses integrate a DNA copy of their genome into host DNA as an obligatory step in their replication cycle. Our work focuses on the molecular mechanism of integration, and in particular on the structure and function of HIV integrase. Integrase is the viral enzyme that carries out the key DNA cutting and joining steps in the integration reaction. The structures of each of the individual domains of HIV-1 integrase have been determined, but their spatial arrangement in the active complex with DNA substrate is unknown. Our recent efforts have been directed at understanding the spatial arrangement of the domains of HIV-1 integrase and how they interact with DNA substrate. We have determined the structure of the catalytic domain of HIV-1 integrase together with the N-terminal domain. An ordered phosphate is found at an identical location close to the active site in all four monomers in the asymmetric unit. This phosphate mimics the phosphate 5' to the scissile phosphate on the viral DNA substrate and provides a first glimpse at the interaction of the active site with DNA. The structure of residues 1-212 of HIV integrase reveals a different dimer interface between the N-terminal domains than was observed for the isolated domain. It also complements the previously determined two-domain structure of the core and C-terminal domains of HIV-1 integrase; superposition of the conserved catalytic core of the two structures results in a plausible full-length HIV integrase dimer. Furthermore, an integrase tetramer formed by crystal lattice exhibits positively charged channels suitable for DNA binding. We are also studying the mechanism of action of a novel cellular factor (BAF) that is involved in blocking self-destructive autointegration of retroviral DNA. BAF is a non-specific DNA binding protein that has the unusual property of bridging together DNA molecules. This results in intermolecular aggregation at high DNA concentration or intramolecular compaction at low DNA concentration. BAF is present in both the cytoplasmic and nuclear compartments. Since it is a major DNA binding activity in the cytoplasm, it can be acquired passively by the preintegration complex without the requirement for a specific recruitment mechanism. Binding of BAF compacts DNA and our current model is that compaction of DNA by BAF within the preintegration complex makes it less available as a target for integration. This model is supported by sedimentation experiments that reveal the salt-stripped complexes to be less condensed than untreated complexes. Reconstitution with BAF restores their condensation and preference for intermolecular integration in parallel. The structure of BAF has been previously determined by both NMR and X-ray crystallography and we have shown that it forms a complex with DNA that is quite unlike that formed by any other known DNA binding protein.